M. Battistin – 1 st September 2009 EN/CV/DC 1 M. Battistin (EN/CV/DC) 1 st September 2009 ATLAS Working Group Thermo-siphon review - Low temperature condensation

M. Battistin – 1 st September 2009 EN/CV/DC 2 Moving the complication out PR thermal screen others 4 liquid lines 4 gas lines condenser liquid tank pneumatic valve manual valve manual valve dummy load Heater BPR pixel 6 X SCT manual valve pneumatic valve Let’s explore the chiller part of this installation: The cycle needs a condensation at low pressure and temperature. C3F8 obliges to condensate between -50°C and -70°C to reach very low pressures, 450 and 200 mbara respectively. Low temperature Chiller UX15 USA15 Surface

M. Battistin – 1 st September 2009 EN/CV/DC 3 The chiller becomes the critical part of the installation

M. Battistin – 1 st September 2009 EN/CV/DC 4 The fridge cascade Primary Water 25°C Cooling source is the primary water coming form the cooling towers: ≈25 °C Two fridge systems in cascade takes the temperature to -70 °C. The first stage uses R11 and the second stage R507a. The second stage cools down a small liquid brine circuit R134a used to condensate the C3F8 in the condenser of the thermo siphon circuit. The condensed C 3 F 8 falls in the liquid receiver due to gravity. First Chiller (R23) Second Chiller (R404a) -30°C-70°C -67°C Condenser Liquid receiver Brine (R134a)

M. Battistin – 1 st September 2009 EN/CV/DC 5 Many possible suppliers The fridge cascade shall be: 1)Fully redundant (2 x 100% or 3 x 60% solutions can be adopted) 2)Highly reliable (proven performances and MTBF) 3)Automatic (the control system will be independent to the thermo siphon one) We will search for fully industrial hand-key solutions realized by reliable company with a proven list of similar installation realized in the past. We are now in the Market Survey (MS) phase. The visit to the companies premises will be part of the selection criteria of the MS. The Invitation to Tender (IT) will be sent to the selected companies only. The long list is: Axima Refrigeration (D), Axima (CH), Angelantoni (I), Froid Seicar (F), Mayekawa Europe (B), Clauger (F), others are going to be selected

M. Battistin – 1 st September 2009 EN/CV/DC 6 Visit to Angelantoni on June 29 th 2009 Clear previous experience on refrigeration for industry: 1)Brine chilling units, package design, air-cooled or water-cooled, refrigerating capacities from 30 to 800 kW, temperature range from 0°C to – 80°C. 2)Refrigerating systems for harmful gas or vapors condensation, to avoid environment pollution, custom-designed according to Customer requirements, capacity range from 30 to kW, temperature range from 0°C to – 80°C. We are waiting (this week) for the first detailed technical and commercial offer for a cooling unit 6 -60°C or -70°C. The budget offer is around 300 k€ 5.0 m 4.0 m 2.0 m ≈300 k€

M. Battistin – 1 st September 2009 EN/CV/DC 7 Space could be a problem

M. Battistin – 1 st September 2009 EN/CV/DC 8 Is there any cheaper solution for a test bench? The liquid temperature increases form the surface to the cavern due to the environmental thermal load. The final (cavern) temperature increases with the mass flow. There is a minimal mass flow required to avoid the temperature increases more than the saturation one. There is a minimal cooling power we need to condensate the minimal flow rate required. Some laboratory chiller can be used in this testing phase to provide the required minimal power. LAUDA JULABO HUBER A minimum requirement is around C

M. Battistin – 1 st September 2009 EN/CV/DC 9 JULABO – FPW91-SL CHF – 6 weeks delivery

M. Battistin – 1 st September 2009 EN/CV/DC 10 HUBER W CHF – 4 weeks delivery

M. Battistin – 1 st September 2009 EN/CV/DC 11 LAUDA ? Waiting for the offer…

M. Battistin – 1 st September 2009 EN/CV/DC 12 Conclusions The low temperature cooling source is the critical part of the thermo siphon installation. It leads the MTBF of the system. We believe that a fully redundant, fully industrial solution shall be adopted. Previous experiences (EN/CV/DC) with industrial contract show how the supplier selection is a very critical point. For the test phase a laboratory chiller seems much more attractive. We are going to launch the Market Survey for the selection of the short list of companies for the supply and installation of this part of the system. The first stage will be for low power (ATLAS IBL design parameters). This will validate the solution for phase II.